Floods in Youghiogheny and Kiskiminetas River Basins Pennsylvania and Maryland
Total Page:16
File Type:pdf, Size:1020Kb
GEOLOGICAL SURVEY CIRCULAR 204 FLOODS IN YOUGHIOGHENY AND KISKIMINETAS RIVER BASINS PENNSYLVANIA AND MARYLAND FREQUENCY AND MAGNITUDE Prepared by Water Resources Division UNITED STATES DEPARTMENT OF THE INTERIOR Oscar L. Chapman, Secretary GEOLOGICAL SURVEY W. E. Wrather, Director GEOLOGICAL SURVEY CffiCULAR 204 FLOODS IN YOUGHIOGHENY AND KISKIMINETAS RIVER BASINS PENNSYLVANIA AND MARYLAND FREQUENCY AND MAGNITUDE Prepared by Water Resources Division Prepared in cooperation with the Pennsylvania State Department of Forests and Waters Washington, D. C.,1952 Free on application to the Geological Survey, Washington 25, D. C. CONTENTS Page Page Introduction............................. 1 Gaging-station records--Continued Method................................... 1 Kiskiminetas River basin--Continued Records Available........................ 1 Conemaugh River at Tunnelton, Pa .•.. 13 Flood-frequency relationships............ 3 Kiskirninetas River at Avonmore ....... 14 Analysis of flood data................. 3 Little Conemaugh River at East Plotting positions..................... 3 Conemaugh........................ 15 Adjustment to base periods............. 3 Blacklick Creek at Blacklick •••••.•.. 15 Test for homogeneity................... 4 Loyalhanna Creek at Kingston ......... 16 Median flood ratios.................... 4 Loyalhanna Creek at New Alexandria •.. 16 Composite flood-frequency graphs....... 4 Youghiogheny River basin: Mean annual flood. • . • . • . • . • . 9 Youghiogheny River at Ohiopyle, Pa ... 17 Computation of comparable means........ 9 Youghiogheny River at Connellsville .. 18 Mean annual flood versus drainage area. 9 Youghiogheny River at Sutersville .... 19 Application of flood data ................ 10 Casselman River at Markleton.......... 20 Stage frequencies ...•••••••..•........... 11 Big Piney Run near Salisbury •...•.•.. 20 Gaging-station records ..•.........•...••• 11 Laurel Hill Creek at Ursina .......... 21 Kiskiminetas River basin: Green Lick Run at Green Lick Stony Creek at Ferndale, Pa .•...•...• 12 Reservoir........................ 22 Conemaugh River at Seward ...•........ 12 ILLUSTRATIONS Page Figure 1. Map showing location of gaging stations included in this report.................... 2 2. Period of record of maximum annual peaks at gaging stations......................... 4 3. Frequency of annual floods, all stations, period 1914-50 ••..• ·...•......•........... 7 4. Frequency of annual floods, stations 4, 6, 9, 1q and 11, period 1914-50............ 8 5. Frequency of annual floods, stations 4, 6, 9, 1q and 11, period 1884-1950.......... 8 6. Variation of mean annual flood with drainage area ..........•.......•............... 10 TABLES Page Table 1. Ratios to mean annual floods, ba~e period 1914-50.................................. 5 2. Ratios to mean annual floods, base period 1884-1950................................ 6 3. Computation of mean annual floods for period 1884-1950, Kiskiminetas and Youghiogheny River basins. • . • . • • . • . • . • . • • . • . • . • . 9 iii FLOODS· IN YOUGIDOGHENY AND KISKIMINETAS RIVER BASINS PENNSYLVANIA AND MARYLAND FREQUENCY AND MAGNITUDE INTRODUCTION by the Pittsburgh suboffice, Max Noecker, engi neer-in-charge, in collaboration with engineers of the Water Resources Division, U. S. Geolog Engineers have long appreciated the fact ical Survey, Washington, D. C. that it is seldom economically sound to design hydraulic structures either for the maximum previous floods or for the computed maximum METHOD probable floods, unless failure of such struc tures involves loss of life or serious property With streamflow records available for gaging damage. Such floods may not occur more often, stations located on various sized watersheds on an average, than once in a hundred or pos in the area, the study resolved itself into sibly even several thousand years. In planning two parts: the determination of the flood sound design for structures, such as culverts frequency characteristics of the region ex or bridge·s, which may involve flood frequencies pressed as a curve in terms of the ratio of a of only 10 to 50 yr, the factors of economic~ given flood to the mean annual flood, and the and useful life of the structure should be con definition of a curve of mean annual flood ex sidered. pressed as a function of the drainage area. With these known factors, it is possible to In the past, records of streamflow have been compute the magnitude-frequency relation of so meager and of such short duration that litt~ floods at any location in the region. reliable information could be obtained on the magnitude and frequency of floods in any par ticular region. As a result it has been the RECORDS AVAILABLE tendency of engineers to overdesign, often causing needless expenditure. All gaging-station records of sufficient length in the two river basins in which the The accumulation of reliable streamflow rec peaks are not seriously affected by artificial ords over a relatively long period now permits regulation were used in this study. These satisfactory studies of flood probabilities in stations all located in Pennsylvania, are: many regions. 1. Stony Creek at Ferndale. This report presents a method, based on 2. Conemaugh River at Seward. actual flood data, for obtaining the magnitude 3. Conemaugh River at Tunnelton. and frequency of floods at any place in the 4. Kiskiminetas River at Avonmore. watersheds of the Youghiogheny and Kiskiminetas 5. Little Conemaugh River at East Conemaug~ River basins, an area comprising 3,651 sq mi 6. Blacklick Creek at Blacklick. in the upper Ohio River basin of western Penn 7. Loyalhanna Creek at Kingston. sylvania and Maryland. 8. Loyalhanna Creek at New Alexandria. 9. Youghiogheny River at Ohiopyle. The data contained in this report were com 10. Youghiogheny River at Connellsville. piled and prepared for publication under a 11. Youghiogheny River at Sutersville. cooperative agreement on water resources in 12. Casselman River at Markleton. vestigations between the Pennsylvania Depart 13. Big Piney Run near Salisbury. ment of Forests and Waters, and the Water Re 14. Laurel Hill Creek at Ursina. sources Division, U. S. Geological Survey. 15. Green Lick Run at Green Lick Reservoir. J. W. Mangan directs the surface-water inves tigations in Pennsylvania for the U. S. Geo The locations of these stations are shown logical Survey. The data presented in this on the map presented as figure 1. report were analyzed, and the text prepared, l 2 FLOODS IN PENNSYLVANIA AND MARYLAND •• •I • PENNSYLVANIA -wEST' ViRGINIA ( ......__ \ ________ I " j PENNSYLVANIA ( : _______ ./' 5 0 20 Miles Figure 1.--Map showing location of gaging stations included in this report. FLOOD-FREQUENCY RELATIONSHIPS 3 Records for the following stations, operated Annual floods Partial-duration-series presently or in the past within this area, have not been used in this report: 1.16 0.5 1.58 1.0 Casselman River at Confluence, Pa., 2.00 1.45 Laurel Hill Creek at Confluence, Pa., 2.5::1: 2.0 Youghiogheny River at Youghiogheny River 5.52 5.0 Dam, Pa., 10.5 10 ~asselman River at Grantsville, Md., 20.5 20 Youghiogheny River at Oakland, Md., 50.5 50 Youghiogheny River at Friendsville, Md. 100.5 100 The first three stations listed were not used because they were affected by backwater Plotting positions for the greater part of the time. The records at the remaining stations were of too short In analyzing the re-lation between frequency duration to obtain satisfac~ory correlations and magnitude of floods, the discharges were with other stations necessary to extend them arranged by number in order of magnitude. to the length of the base period. (.See Ad Appropriate recurrence intervals were then justment to base periods.) assigned to each flood, depending upon the order number, the period of record, and any Whenever possible, gaging-station records additional historical information indicating were supplemented by historical flood data the frequency of recurrence. obtained prior to the systematic collection of stream-gaging records. Recurrence intervals were computed from the formula (li+l)/M, where~ equals number of years Tabulations of the momentary peak stages of record and M equals the order of relative and discharges at gaging stations during each magnitude of the event beginning with the high "water year" of record are presented in the est as one. In the case of a historical flood, last section of this report. The water year N is the number of years during which it is extends from October 1 to September 30. known that the flood was of the order assigned. This formula, adopted by the Geological Survey, is simple to compute, is applicable both to FLOOD-FREQUENCY RELATIONSHIPS annual-flood data and the partial-duration series, and gives results acceptably in con formance with some of the latest theories. Analysis of flood data Annual floods were plotted on a special In developing flood-frequency relationships form2 developed for analysis of flood frequen in this report, the annual-flood method was cies by the theory of extreme values.3 On the used. By this method each flood has been basis of this theory the discharge at a recur treated as an independent event, and ·only the rence interval of 2.33 yr is the same as the maximum momentary discharge occurring in each arithmetical mean of an infinitely long series water year of record was used. of annual floods and is known as the "mean annual flood." As generally is done with rel An objection often raised to the annual atively short